EQUIPMENT CONDITION MONITORING METHOD AND DEVICE THEREOF

摘要

PROBLEM TO BE SOLVED: To reduce data retrieval time searching for data close to new observational data from the entire learning data and to enable abnormality diagnosis of equipment in a short time when carrying out case-based abnormality detection of equipment in a plant etc.SOLUTION: A method includes the steps of: clustering learning data in advance and storing a cluster center and data 106 belonging to a cluster; selecting data close to new observational data from the learning data belonging to a cluster close to the new observational data; preparing a normal model 108 from selected data 107 and determining a threshold 110 by obtaining an abnormal measure 109; and obtaining an abnormal measure from the new observational data and the normal model, and detecting an abnormality of equipment by comparing it with the threshold.